Cross-sectional and prospective associations of early childhood circulating metals with early and mid-childhood cognition in the Project Viva cohort

Background: Relatively little is known about the immediate and prospective neurodevelopmental impacts of joint exposure to multiple metals (i.e., metal mixtures) in early childhood. Objectives: To estimate associations of early childhood (-3 years of age) blood metal concentrations with cognitive test scores at early and mid-childhood (-8 years of age). Methods: We studied children from the Project Viva cohort. We measured erythrocyte concentrations of seven essential (Co, Cu, Mg, Mn, Mo, Se, and Zn) and eight non-essential metals (As, Ba, Cd, Cs, Hg, Pb, Sn, and Sr) in early childhood blood samples. Trained research assistants administered cognitive tests assessing vocabulary, visual-motor ability, memory, and general intelligence (standard deviations: -10 points), in early and mid-childhood. We employed multivariable linear regression to examine associations of individual metals with test scores adjusting for confounders, other concurrently measured metals, and first-trimester maternal blood metals. We also estimated joint associations and explored interaction between metals in mixture analyses. Results: We analyzed 349 children (median whole blood Pb -1 mu g/dL). In cross-sectional analyses, each doubling of Pb was associated with lower visual-motor function (mean difference: -2.43 points, 95% confidence interval (CI): -4.01, -0.86) and receptive vocabulary, i.e., words understood (-1.45 points, 95% CI: -3.26, 0.36). Associations of Pb with mid-childhood cognition were weaker and less precise by comparison. Mg was positively associated with cognition in crosssectional but not prospective analyses, and cross-sectional associations were attenuated in a sensitivity analysis removing adjustment for concurrent metals. We did not observe joint associations nor interactions. Discussion: In this cohort with low blood Pb levels, increased blood Pb was robustly associated with lower cognitive ability in cross-sectional analyses, even after adjustment for prenatal Pb exposure, and regardless of adjustment for metal co-exposures. However, associations with mid-childhood cognition were attenuated and imprecise, suggesting some buffering of Pb neurotoxicity in early life. What this study adds: Relatively few studies have comprehensively separated the effects of neurotoxic metals such as lead (Pb) from preand postnatal co-occurring metals, nor examined persistence of associations across childhood. In a cohort of middle-class children, we found higher early childhood (-3 y) blood Pb was associated with lower scores on cognitive tests, independent of other metals and prenatal blood Pb. However, early childhood Pb was only weakly associated with cognition in mid-childhood (-8 y). Our results suggest the effects of low-level Pb exposure may attenuate over time in some populations, implying the presence of factors that may buffer Pb neurotoxicity in early life.